Hand Recounts in Close Elections: For Future Reference
Just for future reference: we should all agree now that punch-card ballots should be recounted by hand in close elections. That was the position before the election of 2000, and it's the position of all companies I know of that make punch-card ballot voting machines. The consensus of everybody who thinks about such things seriously is that machine counts are approximations, and that hand-recounts are more accurate. That is the procedure that should have been followed in the 2000 election. (And, in fact, by most (but not all) fair ways of counting ballots, Bush would apparently have won if procedure had been followed).
This is the kind of thing that is best settled ahead of time in order to maximize fairness and objectivity, and so that partisan passions don't make us cheat and suddenly decide that a whole new procedure is called for when the old procedure doesn't favor our side.
Changes in technology might make hand recounts less accurate than machines, though. So, what we really need to do I think is:
1. Commit ourselves to pushing to implement the best relevant technology.
2. Make sure there is a paper trail associated with that technology.
3. Make sure it is as tamper-proof as possible.
4. Determine whether hand recounts are more or less accurate than machine counts with that technology.
5. Develop a recount procedure accordingly and ahead of time.
6. Stick with that procedure when push comes to shove.
Another 2000-type debacle in 2008 would, I fear, be disastrous (no matter who comes out on top).
Just for future reference: we should all agree now that punch-card ballots should be recounted by hand in close elections. That was the position before the election of 2000, and it's the position of all companies I know of that make punch-card ballot voting machines. The consensus of everybody who thinks about such things seriously is that machine counts are approximations, and that hand-recounts are more accurate. That is the procedure that should have been followed in the 2000 election. (And, in fact, by most (but not all) fair ways of counting ballots, Bush would apparently have won if procedure had been followed).
This is the kind of thing that is best settled ahead of time in order to maximize fairness and objectivity, and so that partisan passions don't make us cheat and suddenly decide that a whole new procedure is called for when the old procedure doesn't favor our side.
Changes in technology might make hand recounts less accurate than machines, though. So, what we really need to do I think is:
1. Commit ourselves to pushing to implement the best relevant technology.
2. Make sure there is a paper trail associated with that technology.
3. Make sure it is as tamper-proof as possible.
4. Determine whether hand recounts are more or less accurate than machine counts with that technology.
5. Develop a recount procedure accordingly and ahead of time.
6. Stick with that procedure when push comes to shove.
Another 2000-type debacle in 2008 would, I fear, be disastrous (no matter who comes out on top).
posted by Winston Smith at
11:10 AM
3 Comments:
Personally, I think the electronic voting machines are even more of an issue than the punch card ballots as far as close elections go, especially considering how easy it is to tamper with the machine and that there is not a paper trail.
So in my mind, 2 and 3 are the most important things going into the next election. Hearing about machine malfunctions, sketchy election "helpers" tampering with machines, and dead people voting in the last general election makes me feel more than a little worried heading into the next one.
Really, within such narrow ranges as we had in FL, there simply is no fact of the matter has to who won the election, since the very concept of winning an election is not fine grained enough. This is not a crazy po-mo thing, I simply mean that we have no idea about such matters as: whether a dimpled ballot should count, whether real votes require an intention to vote or can be done accidently, whether fraudulant votes discovered long after the fact would invalidate an election, if a box of ballots burns, do we just not count them, or make a statistacally valid assuption, &c. The law and our intuitions are largly silent on these matters, and, when they are not, are infected with political self interest. (See: the GOP on voter qualification, or the Dems on confusion.) Better voting machines will only go so far to solve these kinds of problems, since even a system that was reliable in 999 of a 1000 cases would would taint 150,000 votes.
A far better way of avoiding another 2000 would be to eliminate the Electoral College, which magnifies the importance of close calls in battleground states, as in Florida. It's not impossible that the whole of the popular vote come within the <1000 range, but it's very unlikely. We can't take uncertainty out of quantum-sized political events, but we unhook the cat, as it were.
I happen to agree with most of the above. The Democrats might have stolen the Washington gubernatorial election a few years back, but it was so close anyway. As one fellow put it, If It's Not Close, They Can't Cheat.
I do disagree about the electoral college: I happen to like it as one of the last bastions of federalism, and as a practical matter, if an election ever came in really close, it wouldn't be Florida 2000 times 50, it would be blood and guts and veins in the teeth in every precinct in the nation, and an inauguration the following January would be impossible. Probably even the January after that.
Post a Comment
Subscribe to Post Comments [Atom]
<< Home